The power of photons

Why is light the core of our technology?

The QCi core technology is based upon using light as a versatile tool for many tasks including computing, communications, performing measurements, and generating random numbers. We will explain the specific techniques we use later on, but before we do, it is important to discuss why this is the approach we have chosen. While we all interact with light on a daily basis, we usually don’t think about the underlying physics of light. Light is made of electromagnetic fields which can oscillate and move through space, and is composed of particles known as photons. These particles have interesting quantum mechanical behaviors, but also many other important and useful properties, three very important properties being:

High bandwidth

One advantage of light is that it can carry a lot of information. For example, visible photons of light consist of oscillations that occur hundreds of trillions of times per second. Because these happen so fast, we can reliably distinguish between light with a slightly different frequency, since a small difference will add up quickly. These waves also have very short wavelengths, less than a micrometer, about the size of a bacterium. Even if the frequencies can’t be told apart, a lot of light particles can be placed in a small amount of space and still be told apart. This translates to being able to put a lot of information in a small space and time. Light also has other properties such as polarization, which corresponds to which direction fields in the wave are oscillating along, as well as some more exotic ones which are not important here. In contrast, electronic signals on the other hand typically only have an “on” or “off” state corresponding to two different voltages, and therefore carry much less information. These reasons are why light is often used in telecommunications, but they also make light great for computation and sensing. For sensing the advantage comes from the fact that each particle of light can carry a lot of information about the object we want to understand.

Low (but not too low) energy

A single particle of visible light also carries very little energy, this has two consequences. Firstly, if the photons are used efficiently, the power required to run the computer can be much, much lower than conventional computers. Energy economy is also valuable for sensors. Secondly, a more subtle consequence is that less heat will be produced in a device, making engineering easier. While the energy of a single photon is small, it is still much higher than the energy of thermal vibrations at room temperature, meaning that cooling is not required in either computers or sensors. We can think of this as being in a “Goldilocks” energy range, where the energy for each piece of information can be very low, but not so low that we need to protect it from background heat, which generally costs energy. In contrast, superconducting computers or sensors for example need to be cooled substantially, to protect them. In the case of superconducting qubits to a temperature which is one hundred times colder than deep space is required.

Fast processing

Light travels very fast, in fact from special relativity, information cannot travel faster than the “speed of light” in empty space, which corresponds to going around the Earth more than seven times in a second. Since the light in our devices travels in materials it goes a little slower, about 1/3 to 1/2 of the speed of light in empty space, but this is still very fast. This isn’t actually any faster than the speed at which electrical signals travel. In a way isn’t that surprising, as they are also electromagnetic waves traveling through a metal wire. However, light still has the potential to process information much faster, the reason is that the speed at which the signals can move isn’t usually what limits the rate at which electronics can process information. The true limitations come from switching a transistor moves a lot of electrons around and releases a lot of heat, and running too fast would create too much heat and damage the system. Light not only operates at a lower energy; it also dissipates very little energy as heat, so much faster processing is possible.

Other important properties

There are many other useful properties of light, too many to make an exhaustive list here. For example, the fact that the natural state of light is to be moving means that it is relatively easy to move light particles into contact with each other. From the perspective of computing this is important, as hard optimization problems usually require a high level of interaction and all-to-all connectivity greatly increases the power of hardware solvers for such problems. Also, communication and transfer of information is a key bottleneck in high-performance computing, while light information is constantly in motion.

A key development that allowed electronics to thrive was the ability to miniaturize the components, moving from large vacuum tubes to billions of transistors on a chip. We believe that the same ability is key for technology based on light. Fortunately, light can be made to move around in a material, using the same principles which guide waves in optical fibers. This means that we can also build chips that process light, with electronics integrated into the same chips. While some of our devices may start out on a lab bench, they will eventually end up as compact and robust chip-based devices that can be used by anyone, anywhere. Furthermore, there are many materials that can be used for non-linear optics. Our current platform is lithium niobater because it provides the best optical and optoelectronic properties combined while still being compatible with chip manufacturing processes. As technology advances, new possibilities may become available, too.

The final aspect that we have barely touched on here is that light is fundamentally quantum mechanical and non-linear optics is the key to unlocking these quantum effects. While light is always composed of individual photons, this fact only becomes important for specific kinds of light. We discuss this more in-depth here when we talk about the important aspects of photons that we use.

Challenges of working with light

While light has many attractive properties for sensing, computing, and security applications, there are also important challenges, which we believe we are uniquely equipped to overcome. The biggest challenge is getting photons to interact. Currently, light is widely used in telecommunications, which is an attractive application because it consists of moving information undisturbed rather than processing data while encoded in light. For communication, it is actually better if the particles of light do not interact, as different frequencies can be used to send multiple streams of information through the same optical fiber in a way that is mutually independent of each other. If these channels interacted, then their usefulness for communication would be ruined.

On the other hand, the lack of interactions between light particles which allows large quantities of information to be transmitted through a fiber-optic cable presents a challenge when the light is used instead to compute or otherwise process information. As we discuss here, interactions of some kind between light particles are also needed to make interesting quantum states of light. Computing fundamentally needs interactions of some kind, in classical electronics this could include something like a transistor, where voltage at one electrode can influence the voltage between the other two. Because electric currents can be used to switch other electric currents within transistors, they can be arranged together to build logic gates, where two (or more) voltage inputs can be combined in a predictable way to produce an output voltage, to be used in more logic and build arbitrarily complex systems, including microelectronic computers. In fact a single very simple logic gate, one which outputs a low voltage if and only if both inputs are a high voltage, and a low voltage otherwise (a NAND gate) can be used to build arbitrarily complex logic.

This example with electronics illustrates that we do not necessarily need a large range of interactions between light particles for useful sensing, computing, and security applications. Rather, what we need is to do a few things extremely well and use these operations together for a multitude of applications. This also illuminates why we are targeting a multitude of platforms for our optical technology: the underlying challenges in manipulating particles of light are the same. The unique abilities we are developing for extremely high-quality fabrication of on-chip quantum optic devices will allow us to develop key components which will be used in a versatile way to bring about a revolution in quantum optical devices. By being able to precisely manipulate the behavior of light, eventually, at the single particle level, we will be able to build devices that are currently not possible, including devices that can harness quantum superposition and interference to compute.